Sialylation is a biosynthetic process that involves the addition of sialic acid residues, N- or O-substituted derivatives of neuraminic acid, to the carbohydrate chains, or glycans, of glycoconjugates. Glycoconjugates are molecules that consist of a carbohydrate moiety covalently linked to another chemical moiety, such as a protein, peptide, or lipid, to form conjugates typically classified as glycoproteins, glycopeptides, peptidoglycans, glycolipids, or lipopolysaccharides. Glycoconjugates are used in many biomedical applications, and sialylation is often required for their activity and optimal use. Many therapeutic glycoproteins need to be sialylated, for example, because the sialic acid moieties prevent rapid clearance of a glycoconjugate from the patient's circulation (Morell et al., 1971; Ngantung et al. 2006). Oligosaccharide side-chains, or glycans, are also known to mediate a variety of other glycoprotein functions, including folding, trafficking, stability, and enzyme activity (Harrison and Jarvis, 2006).
Higher eukaryotic production systems, such as mammalian cells, have all of the biosynthetic components required for sialylation, which can be used to produce sialylated glycoconjugates. Lower eukaryotic systems, such as yeast, plant, and insect systems, can also be used to produce recombinant glycoconjugates, often at a lower cost. Lower eukaryotic systems cannot sialylate newly-synthesized glycoconjugates, however, because they lack one or more components of the biosynthetic systems or complexes required for sialylation. These complexes includes many enzymes, such as glycosyltransferases that produce the glycans used as acceptor substrates, sialyltransferases that transfer sialic acid residues from CMP-sialic acids to the acceptor substrates, and enzymes responsible for the production of CMP-sialic acids, which are the donor substrates for glycoconjugate sialylation.
A variety of studies have shown that the inability of yeast, plant, and insect systems to sialylate newly-synthesized glycoconjugates can be addressed by genetic engineering. Each of these host cell systems can be engineered to introduce genes from other organisms that encode the components needed to sialylate newly-synthesized glycoconjugates (Hollister and Jarvis 2001; Hollister et al., 2002; Aumiller et al., 2003; Chang et al., 2005; Hamilton et al., 2006; and Castilho et al., 2010). All of the glycoengineering approaches designed to promote glycoconjugate sialylation so far have also required supplementation of the cellular growth medium with chemicals, such as a sialic acid precursor. The most commonly used precursor has been N-acetylmannosamine (ManNAc), which is converted to ManNAc-6-phosphate (ManNAc-6-P) by intracellular kinases. The resulting ManNAc-6-P can be converted to free sialic acids, which can then be converted to CMP-sialic acids, which are the donor substrates directly required for glycoconjugate sialylation. Media supplementation has its disadvantages, however, as (1) ManNAc is expensive; (2) its addition to the cellular growth medium is inconvenient and increases the risk of cell culture contamination, and (3) the ManNAc supplementation strategy does not necessarily raise the intracellular CMP-sialic acid concentrations to levels that are sufficient to support efficient glycoconjugate sialylation. Earlier studies, for example, have shown that the conversion of ManNAc to ManNAc-6-P by intracellular kinases is a key bottleneck in the efforts to promote sialic acid biosynthesis in lower eukaryotes (Viswanathan et al., 2003).
One popular expression system that is currently limited by the inability to produce sialylated glycoproteins efficiently is the baculovirus-insect cell system, although it has been used to produce a wide variety of other recombinant proteins for biomedical and research applications (Jarvis, 2009; Kost et al., 2005; O'Reilly et al., 1992). All of the established insect cell lines and insects used as hosts for baculovirus-mediated foreign gene expression have less extensive glycoprotein glycan processing capabilities than higher eukaryotes (Geisler and Jarvis, 2009; Harrison and Jarvis, 2006; Jarvis, 2009; Shi and Jarvis, 2007). Recombinant forms of mammalian glycoproteins produced using the baculovirus-insect cell system, therefore, can have functional deficiencies due to the inability of the system to process glycans in a manner similar to that observed in mammalian cells.
The apparent absence of sialic acid metabolism in insect systems was first recognized in 1963 (Warren, 1963), and the notion that insects lack biochemical processes involved in sialic acid synthesis, CMP-sialic acid synthesis and synthesis of sialylated glycoconjugates has been supported by many other studies published over the past 50 years (Marchal et al., 2001; Shi and Jarvis, 2007). A variety of studies have shown that sialic acid synthase, CMP-sialic acid synthetase, and sialyltransferase activities, as well as CMP-sialic acids, are undetectable in lepidopteran insect cell lines, which are commonly used as hosts for baculovirus expression vectors (Aumiller et al., 2003; Hill et al., 2006; Hollister and Jarvis, 2001; Jarvis et al, 2001; Seo et al., 2001; Shi et al., 2007; Tomiya et al., 2001). It is also understood that the lepidopteran insect cell lines and insects used as hosts for baculovirus-mediated foreign gene expression fail to produce recombinant glycoproteins with terminally-sialylated glycans (Geisler and Jarvis, 2009).
Genetic engineering methods now known as “glyco-engineering” have been applied to the baculovirus-insect cell system in recent years to overcome these problems. This approach has involved introducing mammalian genes encoding enzymes involved in glycan processing, sialic acid synthesis, CMP-sialic acid synthesis and glycoconjugate sialylation into insect cell lines or insects in order to improve their endogenous glycoprotein processing capabilities. Glyco-engineering has been accomplished by genetically transforming established insect cell lines (Aumiller et al., 2003; Breitbach and Jarvis, 2001; Hollister and Jarvis, 2001; Hollister et al., 2002; Hollister et al., 1998) or by genetically engineering baculovirus vectors (Hill et al., 2006; Jarvis and Finn, 1996; Jarvis et al., 2001; Lawrence et al., 2001; Seo et al., 2001; Tomiya et al., 2003; Viswanathan et al., 2003). The genetic transformation approach has resulted in transgenic insect cell lines or insects that encode and constitutively express a set of mammalian genes that enable insect cells to produce sialylated glycoproteins. The vector engineering approach has resulted in new baculovirus vectors that not only introduce the gene encoding the recombinant glycoprotein of interest into susceptible cells, but also introduce a set of mammalian genes that enable insect cells to produce sialylated glycoproteins.
Current glyco-engineering methods still pose significant limitations in the ability of insect and other lower eukaryotic systems to produce sialylated glycoconjugates. Specifically, current glyco-engineering methods are limited in the production and processing of sialic acid precursors that are required to produce sialylated glycoconjugates (Viswanathan et al., 2003). New methods which offer cheaper, simpler, and more effective ways to engineer insect and other eukaryotic cells to produce sialylated glycoconjugates are therefore needed. One promising approach would be to engineer these lower eukaryotic systems to produce large, intracellular pools of CMP-sialic acids which are required for efficient sialylation of glycoconjugates, in a variety of eukaryotic expression systems, including fungal, plant, mammalian, and insect cell-based systems.
FIG. 1 sets forth an illustration showing the metabolism of sialic acid in vertebrates. Key differences in the N-glycan processing capabilities of insect and mammalian cells are illustrated in FIG. 2. Terminal glycosyltransferases, such as N-acetylglucosaminyltransferases, galactosyltransferases, and sialyltransferases, are often absent or present only at insufficient functional levels in insect systems. Differences in their protein glycosylation pathways also reflect the inability of insect cell lines and insects to synthesize and transport sialic acids and CMP-sialic acids, which are needed as donor substrates for N- and O-glycan sialylation by sialyltransferases. These differences provided the rationale for using mammalian genes encoding glycosyltransferases and other enzymes involved in sialic acid and CMP-sialic acid biosynthesis as the targets for earlier glyco-engineering efforts as shown in FIGS. 3 and 4.
The methods described herein greatly enhance the ability of genetically-modified host cell systems to facilitate the production of sialylated glycoproteins, by eliminating the need to supplement the cell culture media with expensive metabolic precursors, such as N-acetylmannosamine, needed to promote the efficient sialylation of recombinant proteins. In this study, genetically-modified insect cells were made which produce sufficient amounts of the required precursor. The cells were engineered to contain a variety of nucleic acids encoding polypeptides derived from mammalian sources needed to promote the sialylation of recombinant glycoproteins, plus a nucleic acid encoding E. coli N-acetylglucosamine-6-phosphate 2′-epimerase (GNPE), which normally functions in bacterial sialic acid degradation. Under normal conditions, these cells have the product, but not the substrate for this enzyme. Modified cells that comprise a nucleic acid encoding a GNPE that is expressed at a sufficient level could drive the reaction in reverse, initiating sialic acid biosynthesis in the absence of media supplementation. The modified insect cells efficiently produced sialic acid, CMP-sialic acid, and sialylated recombinant N-glycoproteins even in growth media without N-acetylmannosamine. This approach is not limited to insect cells, can be adapted to a variety of other eukaryotic host cell systems. The general scheme is illustrated in FIG. 5.